An Exploration of Fetish Social Networks and Communities
Total Page:16
File Type:pdf, Size:1020Kb
View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Bournemouth University Research Online An exploration of fetish social networks and communities. Damien Fay, Hamed Haddadi, Michael C. Seto Han Wang, and Christoph Kling Department of Computing, Bournemouth, UK School of Elec. Eng. and Comp. Sci., Queen Mary University, UK The Royal's Institute of Mental Health Research, Ottawa, Canada National University of Ireland, Galway The Leibnitz Institute for the social sciences, Cologne, Germany {dfay.bournemouth.ac.uk} Abstract. Online Social Networks (OSNs) provide a venue for virtual interactions and relationships between individuals. In some communities, OSNs also facilitate arranging offline meetings and relationships. FetLife, the worlds largest anonymous social network for the BDSM, fetish and kink communities, provides a unique example of an OSN that serves as an interaction space, community organizing tool, and sexual market. In this paper, we present a first look at the characteristics of European members of Fetlife, comprising 504,416 individual nodes with 1,912,196 connections. We looked at user characteristics in terms of gender, sexual orientation, and preferred role. We further examined the topological and structural properties of groups, as well as the type of interactions and relations between their members. Our results suggest there are important differences between the FetLife community and conventional OSNs. Keywords: Social network properties, sexuality, topic modelling 1 Introduction Social interaction is motivated at the individual level in need for power, pres- tige and approval [22] which are expressed in modern life in activities such as business, friendship/emotional learning exchange, and knowledge exchange; and from an evolutionary perspective the need to seek a mate. This latter function of a social network is known as the sexual market and every social network has a secondary function as a sexual market, although disaggregating this function from others can be challenging [12]. In the last decade, Online Social Networks (OSNs) have become a focal point of the web and the most popular activity of individuals online. There are a large number of popular OSNs and a large body of research focuses on a variety of OSNs. Despite a large number of papers on analysis of large scale OSNs [14, 2], and a large number of social science papers on social relationships, sexuality and orientations [15][10], there have not been 2 Fay, Haddadi, Seto, Wang, Kling any academic papers which have examined online social networks focused on variations in sexual orientations and interests. In this paper, we take a first look at the anonymised profiles of the European users of the most popular fetish website, and ask if the characteristics of the network are different from those of a conventional OSN. This is a rich dataset of over half a million users and captures patterns of traditionally secret interests and behaviours. We do so by comparing the topological characteristics with those of popular social networks as reported by Mislove et al. [14]. We choose this online fetish network as it is oriented towards friendships, social groups, and arranging events, where the social is primary and sexual market is secondary but explicitly included (unlike, say, Facebook or other non-dating OSNs). It is important to social scientists and psychologists to understand whether a social network is also present or not required. As FetLife reveals sexuality in a social context it allows us to understand sexual networks in a way that dating sites such as Tinder, Grindr etc might not allow; this is also vital for creating models for the spread of sexually transmitted infections [17]. We use our large dataset to assess the properties of these multi-relationship networks, where a user can have a number of different types of relationships with others.1 We base our analysis of the structure of the graph on work by Laumann et al. [12], who use self reports and assess individuals' roles and eco- nomic factors in sex markets, using four neighbourhoods in Chicago and high- light the role of brokers and third parties in this exchange. Our dataset uses the largest broker out there, the world's most popular fetish site, as a benchmark for analysis of the online version of this market. Understanding the nature of the interactions is also important for real and cyber crime investigations, as the privacy and safety of users could also be compromised by malicious users of such websites.2 2 Online Fetish Networks We collected our data from FetLife,3 the most popular Social Network for the BDSM, Fetish, and kink communities, with millions of users worldwide. The fetish community has grown rapidly in recent years and now consists of a diverse collection of people whose interests cover a broad spectrum including, fashion, burlesque, a nightclub scene, particular types of music and of course a focus on sexual experimentation. As in Facebook, the interaction of the community is both real-world and virtual with a large collection of real-world events attended by members; contrary to expectations, FetLife it is not a paid dating website. For example, there is no \search" functionality within the website for specific types of members, e.g., based on interests, or over user information fields (height, weight, age, location, fetish commonalities, other personal information). However 1 In the interest of space and scientific focus, we encourage the readers to see [13, 16] for a description of different types of fetish relationships. 2 http://sexandthe405.com/fetlife-is-not-safe-for-users/ 3 https://fetlife.com/ An exploration of fetish social networks and communities. 3 the site is used as a bootstrapping mechanism for social events, workshops, and parties which are organised regionally. Members create a personal profile, similar to most OSNs, specify their gender, age, role, orientation, and list the fetishes they are interested in or are curious about. The users are organised into tens of thousands of groups, and thousands of events are arranged annually through the website. Users pay particular attention to the experience of the group members and event organisers and hence these individuals play a central role in the community. In essence, FetLife is a niche OSN. BDSM is a sexual interest or subculture attractive to a minority [18]. What makes FetLife unique particularly interesting for OSN analysts is that this website observes sexual interaction (present in dating websites, absent in typical social networks such as Facebook) but in the presence of a social context (absent in dating websites). 3 Data collection We collected our data from the European members of FetLife during the early months of 2014. The data includes anonymized (at the time of collection) user IDs, relationship types, and number of friends. In order to comply with the web- site policy and ethics approval requirements, we did not crawl any names, details of friends, pictures, posts, or other personally identifiable information available on the site. Since it is mandatory for users to be a member of a single geographic area (usually county/borough level depending on the population density), our crawler used the location area codes of the website as its seed and we collected the mentioned details about every single individual in the European section of the website. Overall, there are 504,416 individual nodes in our dataset, with 1,912,196 connections. The main connected component is comprised of just over 156K nodes, and the rest of the users are mainly isolated or small groups of maximum size 20. At the time of collection, there were 35,153 groups in the dataset, with just over 26k single nodes. Although this is a sample of the population and only captures the individuals who chose to be on a fetish OSN, this data is more inclusive and less biased than the offline club members or those who self-identify for sample surveys in existing literature [4, 18]. The perceived anonymity online and low (essentially zero) cost of entry into Fetlife means more individuals might be active online than joining actual clubs, going to local BDSM themed parties or self-identifying to researchers at universities. 4 Demographic analysis In this section we document the demographics of the fetish network such as gender, sexual preference, and connections. The identity acronyms are defined as follows: M = cis male; F = cis female; TV = transvestite; TS = transsex- ual, which can be further distinguished into male-to-female transsexuals (MtF or trans females) and female-to-male transsexuals (FtM or trans males); Ka- jira/Kajiru are slave girl/boy; I = intersex, B = butch, Fem = Femme. If not 4 Fay, Haddadi, Seto, Wang, Kling 80 70 All users 60 Degree >1 Degree >5 50 % 40 30 20 10 0 ns M F TV MtF FtM TG GF GQ IS B FEM Gender Fig. 1: Distribution of genders for all users, users with > 1 friends, and > 5 friends. otherwise stated, Trans = trans females and TVs. GF = gender fluid and GQ = gender queer, referring to persons who do not identify as male or female or see themselves as having aspects of both genders. We first look at the gender demographics of the users as a whole. As mentioned previously, there are larger number of users with no friends than would otherwise be expected. Figure 1 shows the distribution of user gender for all users. When the singletons have been removed, the gender distribution changes drastically; most of those with few or no friends are male (Figure 3 shows that in addition they tend to be heterosexual males). When we have taken out those with fewer than 5 friends then the gender distribution is quite even with (non-cis) 54% male, 40.5% female and other (cis) genders making up the remainder.